I. Introduction
This is the final installment of our five-part series on the Stolt-Nielsen oral argument. It addresses the fourth issue identified in Part I (here): what the import of the agreements’ silence is or should be. It assumes the Court reaches the merits; as explained in Parts III and IV (here and here), the United States Supreme Court may take another “pass” on the question presented (the first pass was taken in Bazzle), and hold that the predicate for granting certiorari was not established because the arbitrators ruled that the agreement was not silent on class arbitration.
Part I identified two loose “coalitions” of Justices – the “Breyer Coalition” consisting of Associate Justices John Paul Stevens, Stephen G. Breyer, and Ruth Bader Ginsburg, and the “Roberts Coalition,” consisting of Chief Justice John G. Roberts and Associate Justices Antonin G. Scalia and Samuel A. Alito, Jr. The Breyer Coalition appears to be leaning toward either taking a pass or affirming the decision of the United States Court of Appeals for the Second Circuit, which upheld the arbitrators’ award imposing class arbitration, while the Roberts Coalition appears to be leaning toward reversal. We explore the import of the charter-party agreements’ silence on class arbitration from the standpoint of both coalitions.
II. Breyer Coalition
There are at least four outcomes that the Breyer Coalition might consider acceptable: (a) a ruling that that the arbitrators acted within the scope of their authority in imposing class arbitration on the parties because the parties submitted that issue to arbitration (see Part II, here); (b) a ruling that the arbitrators acted within the scope of their authority by ruling that the broad arbitration clause, coupled with the parties’ failure to prohibit class arbitration procedures, evidenced the parties’ consent to class action arbitration; (c) a ruling that class arbitration may be imposed by arbitrators if state arbitration law permits courts or arbitrators to impose it; or (d) a ruling that the predicate for granting certiorari was not met because the arbitrators found that the contracts were not silent on class arbitration (see Part III, here). In the event that the Breyer Coalition commands a majority – it appears to have the support of at least three Associate Justices – and the Court reaches the merits of the question presented, we believe that the Court will rule that arbitrators do not offend the Federal Arbitration Act by imposing class arbitration in the face of the agreement’s silence on that point.
There would be some fundamental problems with such a ruling. First, it would not be grounded in the Court’s prior jurisprudence providing that the purpose of the Federal Arbitration Act is to enforce arbitration agreements as written and according to their terms, even if the parties’ freely-bargained-for dispute resolution method may lead to inefficiency in the form of multiple arbitration proceedings. See, e.g., Volt Information Sciences, Inc. v. Board of Trustees of Leland Stanford Junior Univ., 489 U.S. 468, 479 (1989); Dean Witter Reynolds Inc. v. Byrd, 470 U.S. 213, 220 (1985). For the reasons explained in Section IV., below, enforcing the Stolt-Nielsen charter parties as written means finding that consent to class arbitration cannot be implied.
Second, allowing arbitrators to find consent to class arbitration based solely on the broad scope of an arbitration agreement allow arbitrators to make random, subjective, and virtually unreviewable public-policy-based judgments about whether class arbitration should be imposed in a given case. Those judgments should either be made by the parties or Congress.
Third, because arbitrators will likely continue to impose class arbitration frequently, a Court ruling encouraging them to do just that would effectively elevate a judicial public policy preference for efficient (and perhaps fairer) private dispute resolution above the pro-enforcement policies of the Federal Arbitration Act. While this may be a laudable goal, it is one for Congress to implement, not the Court. The Court’s role is limited to implementing the only public policy the Act was designed to foster: promoting the enforcement of private agreements to arbitrate according to their terms.
Fourth, potentially implicit in such a ruling would be the application of a default rule that class or consolidated arbitration is permitted when the parties’ contracts are silent on that score. But the Court’s prior jurisprudence provides that the sole source of default rules of contract construction or interpretation is state law applicable to contracts generally. No such rule exists in New York general contract law or federal maritime law (see Section IV), which means that the Court would have to create a federal one. It is unclear to us on what basis such a rule could legitimately be formulated, and doing so would be inconsistent with — and potentially throw into disarray — established rules of Federal Arbitration Act preemption.
The Breyer Coalition might avoid some of these problems by finding that state arbitration law permitting courts or arbitrators to impose class arbitration is not preempted by the Federal Arbitration Act, which is silent on whether courts or arbitrators can compel class or consolidated arbitration in the absence of the parties’ express or implied agreement. But such a ruling would beg the question whether compelling class or consolidated arbitration is consistent with enforcing as written each individual arbitration agreement between each individual party, something that the Federal Arbitration Act requires. Additionally, it would leave to the vagaries of state law the question whether class or consolidated arbitration may be imposed, and it might breed litigation over choice-of-law questions.
If the Breyer Coalition commands a majority, these problems may persuade that majority to dispose of the case on a more narrow ground, such as a ruling that the parties submitted the issue of class arbitration to the arbitrators, or that the predicate for granting certiorari simply was not met. For the reasons set forth in Parts II and IV (here and here), we believe that outcome would be unfortunate.
III. Roberts Coalition
The Roberts coalition appears to be willing to address the merits of the Stolt-Nielsen question, presumably because these Justices are not satisfied with the Bazzle status quo under which courts have effectively ceded to arbitrators a great deal of power to impose class or consolidated arbitration without meaningful review by courts. These Justices, we suspect, are concerned that the practical effect of Bazzle is at odds with what the Court has repeatedly said is the principal purpose of the Federal Arbitration Act.
We think the Roberts Coalition would favor a return to the pre-Bazzle regime under which class or consolidated arbitration was the exception, not the rule. Relying on the “enforce as written” principle, the majority of pre-Bazzle decisions held that courts may not impose class or consolidated arbitration in the absence of the parties’ express (or at least implied) agreement. See, e.g., Glencore, Ltd. v. Schnitzer Steel Products, 189 F.2d 264 (2d Cir. 1999); United Kingdom v. Boeing Co., 998 F.2d 68 (2d Cir. 1993); Champ v. Siegal Trading Co., 55 F.3d 269 (7th Cir. 1995). These courts eschew the notion that the efficiency that might result from imposing consolidated or class arbitration justifies doing so in the absence of the parties’ consent.
If the Roberts Coalition garners the support of five Justices – it appears to have at least three votes, and might have the support of Associate Justices Anthony M. Kennedy and Clarence Thomas – we believe that the Court will hold that courts or arbitrators cannot impose class or consolidated arbitration in the absence of the parties’ express or implied agreement. And in Stolt-Nielsen that would mean the arbitrators exceeded their powers by imposing class arbitration on the Stolt-Nielsen entities.
In Section IV., below, we briefly explain some of the reasons why we believe that the Roberts Coalition would likely conclude that enforcing the Stolt-Nielsen charter-party contracts as written requires separate arbitration proceedings between each Stolt-Nielsen entity and AnimalFeeds under each individual charter party between them, and between each Stolt-Nielsen entity and each putative class member.
IV. Why Courts and Arbitrators Should Not Impose Class or Consolidated Arbitration when the Parties’ Agreements are Silent on that Point
There are two key provisions in the charter parties — and for that matter, in most arbitration agreements — which, if enforced as written, strongly suggest that the parties never authorized class or consolidated arbitration: (a) the provisions of the agreement identifying the parties to the contract; and (b) the provisions governing arbitrator selection.
A. Provisions Identifying the Parties
The provisions denoting the parties to the Agreement are fundamental in Federal Arbitration Act cases because they determine who is obligated to whom and who is subject to the arbitrators’ jurisdiction. For example, courts will not hesitate to vacate awards that purport to bind or confer a benefit in favor of a non-party to the agreement. See, e.g., Nationwide Mutual Ins. Co. v. Home Ins. Co., 330 F.3d 843, 846-47 (6th Cir. 2003); Orion Shipping & Trading Co., Inc. v. Eastern States Petroleum Corporation Of Panama, S.A., 312 F.2d 299, 300-01 (2d Cir. 1963). And courts will not enforce arbitration agreements on behalf of, or against, non-signatories, unless state law applicable to ordinary contracts permits such enforcement. See Arthur Andersen LLP v. Carlisle, 556 U.S. ___, 129 S. Ct. 1896, 1901-02 (2009); Ross v. American Express Co., 547 F.3d 137, 143 & n.3 (2d Cir. 2008); Ross v. American Express Co., 478 F.3d 96, 99 (2d Cir. 2007); Astra Oil Co. v. Rover Navigation, Ltd., 344 F.3d 276, 279-80 & n.2 (2d Cir. 2003); Thomson-CSF, S.A. v. American Arbitration Assoc., 64 F.3d 773, 776-80 (2d Cir. 1995).
There is no reason that disputes concerning class arbitration should be decided under a different set of rules. When parties A and B agree to arbitrate their disputes arising out of or relating to their contract, and do not provide for arbitration with parties C, D or E, or with respect to identical contracts that party A may have with parties C, D or E, parties A and B have evidenced an intent to limit the scope of their private dispute resolution mechanism to disputes between them. It may be that parties A and B also intended to arbitrate disputes between them arising under contracts relating to the contract containing the arbitration clause (for example, a separate agreement to terminate or rescind the contract containing the arbitration clause), but they have not expressed an intent to arbitrate with C, D or E disputes that may relate in some way to the contract between A and B.
Bazzle arguably allows arbitrators to impose class arbitration where party A is a party to identical, but separate, contracts between B, C and D. Looking at each contract individually, A agreed to arbitrate in a single proceeding only with B, C or D, not all three. But, where B, C or D all want to arbitrate in a single proceeding, and where A has selected the same arbitrator for its separate arbitrations with B, C, and D, that arbitrator, under Bazzle, can consolidate the proceedings or order class arbitration. In other words, it allows the arbitrator to look past the reality that A entered into separate contracts with B, C, and D, and transform B, C and D’s collective desire to arbitrate in a single proceeding, and A’s appointment of the same arbitrator for all three proceedings, into consent by A to arbitrate with B, C, D and any putative class members in a single proceeding. To justify escaping the reality that A never entered into a single contract to which B, C, D and the putative class members were all parties, and that A does not consent to class or consolidated arbitration, Bazzle says that A, by granting broad authority to the arbitrator, effectively consented to the arbitrator imposing consolidated or class arbitration as a matter of “procedure.”
We think the Roberts Coalition will conclude that such reasoning is not permitted by the “enforce as written” principle. But even if the Roberts Coalition stops short of disapproving the plurality’s logic in Bazzle, the facts in Stolt-Nielsen are fundamentally different. The Stolt-Nielsen defendants are not one but several separate entities, and, to our knowledge, there are separate, materially indentical contracts between each Stolt-Nielsen entity and AnimalFeeds. There are also separate contracts between individual Stolt-Nielsen entities and individual, putative class members, but which are not identical to those the Stolt-Nielsen entities entered into with AnimalFeeds. And as far we know: (a) the Stolt-Nielsen entities have not agreed to arbitrate together as a group against any individual class member, let alone with all of them; and (b) each Stolt-Nielsen entity does not have a contractual relationship with each and every putative class member (other than AnimalFeeds). For the Court to imply consent to class arbitration of the type sought by AnimalFeeds, at a minimum, each Stolt-Nielsen entity must have agreed to arbitrate with each and every member of the putative class. But that does not appear to be the case.
1. Relevance of New York State Law Applicable to Contracts Generally
New York State law applicable to contracts generally does not provide any support for implying consent to class arbitration. First, under New York law, a contract’s silence on a given point ordinarily does not mean that the contract is ambiguous, and thus subject to interpretation and construction. To ascertain whether a contract is ambiguous, courts focus on what is said, not what is omitted:
An omission or mistake in a contract does not constitute an ambiguity [and]. . . the question of whether an ambiguity exists must be ascertained from the face of an agreement without regard to extrinsic evidence.
Reiss v. Financial Performance Corp., 97 N.Y.2d 195, 199 (2001) (citations and quotations omitted). This rule should preclude a finding that the agreements’ silence enables the court (or enabled the arbitration panel) to rule that the agreements can be reasonably construed as permitting class or consolidated arbitration.
Second, where, as here, parties omit to provide for a contractual contingency, New York law generally prohibits courts from adding implied terms to the contract, and always forbids courts from departing from the plain meaning of the words used:
Even where a [contractual] contingency has been omitted, we will not necessarily imply a term since courts may not by construction add or excise terms, nor distort the meaning of those used and thereby make a new contract for the parties under the guise of interpreting the writing.
Reiss, 97 N.Y.2d at 199 (citations and quotations omitted); see also Vermont Teddy Bear Co. v. 538 Madison Realty Co., 1 N.Y. 3d 470, 475 (2004). While implying a term may be appropriate under certain limited circumstances, New York courts “will not imply a term where the circumstances surrounding the formation of the contract indicate that the parties, when the contract was made, must have foreseen the contingency at issue and the agreement can be enforced according to its terms.” 97 N.Y.2d at 199.
Implying into the parties’ agreements consent to class or consolidated arbitration would violate this principle. During the pre-Bazzle period the contract were entered into, the majority rule under the Federal Arbitration Act prohibited courts from compelling class or consolidated arbitration in the face of an agreement’s silence. These cases – including two in the Second Circuit — made clear that parties had to expressly agree to consolidated or class arbitration. And while the notion of class arbitration in the maritime context was unheard of, attempts to compel consolidated arbitration were not. The parties were therefore on notice that, if they wished to arbitrate on a class or consolidated basis, they would have to provide for that contingency in their agreement. But they did not.
2. Relevance of Federal Maritime Law
Federal maritime law must also be considered because the arbitration clauses are contained in charter-party contracts, which are maritime contracts governed by a combination of federal maritime and state law. But to the extent that federal maritime law is relevant, it provides no support for implying consent to class arbitration.
Federal maritime law requires courts to interpret maritime contracts in light of custom and usage. See Stolt-Nielsen, 435 F. Supp 2d at 385-86 (citing authority). The district court held that the arbitrators manifestly disregarded this principle because the undisputed evidence of custom and usage showed that there had never before been a class arbitration arising out of charter-party agreements. To the extent it finds that federal maritime law is relevant, the Supreme Court should find that its application would preclude the Court from implying consent to class arbitration.
B. Imposing Class Arbitration would Deny the Parties the Benefit of their Freely-Bargained-For Arbitrator Selection Provisions
Perhaps no provisions in arbitration agreements are more sacrosanct than those governing arbitrator selection. Circuit Judge Posner put it well when he said: “Selection of the decision maker by or with the consent of the parties is the cornerstone of the arbitral process.” Lefkovitz v. Wagner, 395 F.3d 773, 780 (7th Cir.), cert. denied, 546 U.S. 812 (2005).
Chapters 1 and 2 of the Federal Arbitration Act expressly provide for court enforcement of arbitrator selection provisions. Federal Arbitration Act § 5 provides that, “[i]f in the agreement provision be made for a method of naming or appointing an arbitrator or arbitrators or an umpire, such method shall be followed. . . .” Article V(1)(d) of the Convention on the Recognition and Enforcement of Foreign Arbitral Awards ( here) provides a defense to recognition and enforcement of an award where “[t]he composition of the arbitral authority or the arbitral procedure was not in accordance with the agreement of the parties.” Courts will not hesitate to vacate domestic or non-domestic awards if arbitrator selection procedures were not followed. See, e.g., Encyclopaedia Universalis S.A. v. Encyclopaedia Brittanica, Inc., 403 F.3d 85, 91-92 (2d Cir. 2005); Cargill Rice, Inc. v. Empresa Nicaraguense Dealimentos Basicos, 25 F.3d 223, 226 (4th Cir. 1994); Avis Rent A Car Sys., Inc. v. Garage Employees Union, 791 F.2d 22, 25 (2d Cir. 1986).
The arbitrators have effectively determined that the parties did not agree that they would have the benefit of selecting different arbitrators for each dispute that might arise under each individual arbitration agreement. As Chief Justice William H. Rehnquist said in his dissent in Bazzle (which Associate Justice Anthony M. Kennedy joined):
[P]etitioner had the contractual right to choose an arbitrator for each dispute with the other 3,734 individual class members, and this right was denied when the same arbitrator was foisted upon petitioner to resolve those claims as well. Petitioner may well have chosen different arbitrators for some or all of these other disputes; indeed, it would have been reasonable for petitioner to do so, in order to avoid concentrating all of the risk of substantial damages awards in the hands of a single arbitrator.
539 U.S. at 458-59.
Although Stolt-Nielsen does not press the issue before the Court, the conflicting arbitrator qualifications specified by the two arbitration agreements raises, what we believe can be a vitally important issue in class arbitration cases. The arbitrator selection provisions of the Vegoilvoy charter-party arbitration agreements between Stolt-Nielsen and AnimalFeeds, provide that each party shall appoint “an arbitrator, who shall be a merchant, broker or individual experienced in the shipping business; the two thus chosen, if they cannot agree, shall nominate a third arbitrator who shall be an Admiralty lawyer.” The arbitration agreements governing certain transactions between Stolt-Nielsen and putative class members (which were not parties to the Second Circuit appeal), are contained in the Asbatankvoy charter-party agreement, and provide that arbitration shall proceed “before a board of three persons, consisting of one arbitrator to be appointed by the Owner, one by the Charterer, and one by the two so chosen.” But the arbitrators in Stolt-Nielsen appear to have effectively decided that, for the merits phase of a class arbitration, both parties must either: (a) appoint merchants, brokers or other individuals experienced in the shipping business as party-appointed arbitrators, who will, in turn, select an admiralty lawyer as a third arbitrator; or (b) appoint whomever they please as party-appointed arbitrators, who will, in turn, appoint whomever they please as the third arbitrator. But no matter what the arbitrators might have intended, results (a) and (b) are both in derogation of the unambiguous terms of one of the two form arbitration clauses.
We look forward to a decision in this fascinating and very important case, and shall keep readers apprised of developments as and when they occur. . . .
